Menu
Home
Log in / Register
 
Home arrow Computer Science arrow Social Computing, Behavioral-Cultural Modeling and Prediction
< Prev   CONTENTS   Next >

5 Conclusion

This paper investigated the problem of predicting which author will adopt the novel topics, and tested two algorithms using the DBLP dataset. To solve this problem, the basis is to find the inherent influence network from past observations. The first approach NetInf* is based on the direct observations of past following behavior on topic adoption. It has the shortcoming that the prediction performance is much dependent on the amount of collected past cascades and whether the novel topic cascade is relevant to these past

Fig. 4. Feature Weight Distribution

Fig. 5. Feature Weight Distribution

cascades. Though all topic terms selected are in the same area (machine learning), all these terms are unique. Each topic can have its own unique cascade which could be totally different from past ones. Alternatively, HetNetInf defines the author's following behavior based on the factors affecting the adoption process, such as social connections and topic popularity. Given the adoption information of the author's social connection peers and the topic popularity, it can estimate the probability the author adopting the topic, even the topic is totally new and irrelevant to past ones. The experiment using DBLP shows that HetNetInf outperforms NetInf* in predicting the novel topic adoption. Furthermore, HetNetInf provides information on individual author's preference on the influence factors.

References

1. Gomez-Rodriguez, M., Balduzzi, D., Scho¨lkopf, B.: Uncovering the temporal dynamics of diffusion networks. In: International Conference on Machine Learning, pp. 561–568 (2011)

2. Gomez-Rodriguez, M., Leskovec, J., Krause, A.: Inferring networks of diffusion and influence. ACM Trans. Knowl. Discov. Data 5(4), 21:1–21:37 (2012)

3. Ji, M., Han, J., Danilevsky, M.: Ranking-based classification of heterogeneous information networks. In: Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2011, pp. 1298–1306. ACM, New York (2011)

4. Kempe, D., Kleinberg, J., Tardos, E.: Maximizing the spread of influence through a social network. In: Proceedings of the Ninth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2003, pp. 137–146. ACM, New York (2003)

5. Myers, S.A., Leskovec, J.: On the convexity of latent social network inference. In: Advances in Neural Information Processing Systems, pp. 1741–1749 (2010)

6. Sun, Y., Barber, R., Gupta, M., Aggarwal, C.C., Han, J.: Co-author relationship prediction in heterogeneous bibliographic networks. In: Proceedings of the 2011 International Conference on Advances in Social Networks Analysis and Mining, ASONAM 2011, pp. 121–128. IEEE Computer Society, Washington, DC (2011)

7. Sun, Y., Han, J., Aggarwal, C.C., Chawla, N.V.: When will it happen?: relationship prediction in heterogeneous information networks. In: Proceedings of the Fifth ACM International Conference on Web Search and Data Mining, WSDM 2012, pp. 663–672. ACM, New York (2012)

8. Sun, Y., Han, J., Yan, X., Yu, P.S., Wu, T.: Pathsim: Meta path-based top-k similarity search in heterogeneous information networks. PVLDB 4(11), 992–1003 (2011)

 
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Philosophy
Political science
Psychology
Religion
Sociology
Travel